RESEARCH · ALON GORENSHTEIN, MD
Language models and brain–computer interfaces
Alon Gorenshtein’s brain–computer interface research examines the role of language models in assistive communication and control. His published systematic review distinguishes proposed capabilities from results actually demonstrated in studies.
Research question
Where do language models enter a brain–computer interface pipeline, and what evidence supports each form of integration?
Methods and findings
The review examined eleven studies using a PRISMA-based approach. It classified five ways of combining language models and BCIs: completing text, correcting decoded output, expanding intended messages, generating interfaces, and providing affective support. The analysis considered hardware, decoding, prompting, outcomes, and reporting of latency.
The studies reported promising task-level results, but none enrolled people with motor impairments. Reporting of end-to-end delay and failure modes was limited. The review’s main outputs are a taxonomy and a reporting checklist; clinical benefit for the intended patient population remains an open question.
Authorship
First author of the systematic review, with Mahmud Omar, Yiftach Barash, Girish N. Nadkarni, and Eyal Klang.
Publications and sources
- Large language models integrated into brain–computer interfaces for communication and control: a systematic review
Alon Gorenshtein, Mahmud Omar, Yiftach Barash, Girish N. Nadkarni, Eyal Klang. Biomedical Physics & Engineering Express (2026).
DOI: 10.1088/2057-1976/ae737b
Research summary updated October 26, 2026.